AI and personalisation: Beyond the hype
Of all the buzzwords creating a stir in the gambling industry, personalisation is both one of the most prevalent and the hardest to define.
From targeted advertising to personalised betting offers, it has been used as a catch-all term for new trends and technologies that could help solve the industry’s problems: reducing churn, enhancing the customer experience and making sense of an unwieldy mass of online data.
But for experts who have worked in artificial intelligence (AI) and personalisation since long before the current hype, the personalisation of betting and gaming products is simply a logical progression.
The next step in gaming and media’s evolution
“We look at this as the next step in the gaming and media industry,” says Chris Reynolds, CEO and co-founder of Epoxy, a provider of AI- and machine learning (ML)-driven personalisation solutions in the betting and igaming industry.
“Whether it’s TV, whether it’s music, whether it’s ecommerce, the application of advanced technology to create greater insights into user behaviour and to streamline workflows and derive key metrics like customer lifetime value – these technologies have been applied and in place for a long period of time.”
When it comes to the personalisation of websites and apps, there’s also evidence that bettors have come to expect it: research carried out by Betting Hero in 2023 found that 80% of customers rated personalised offers and bets as “valuable” or “very valuable” and that 75% found non-personalised apps difficult to navigate.
Adoption issues
Nevertheless, the take-up in the industry has remained slow, with operators struggling to overcome the technical and organisational hurdles to embracing AI-driven personalisation.
“I think it is an issue of prioritisation, organisational structure and technical expertise,” says Epoxy co-founder Jason Angelides. “The industry is acutely aware of the value proposition, but implementation of personalisation at scale is not an easy undertaking and there are many existing priorities that take precedence.”
Angelides also points out that adopting AI can involve rethinking some deeply entrenched ideas and systems, such as the siloing of sportsbooks and casinos. “A large majority of organisations are fragmented both operationally and technically,” he says.
One key question for many operators is how to identify where AI and personalisation are most needed and actually add value to both the operator and the consumer. Experts say this starts at the very first stage of the customer journey and can impact everything from retention to harm reduction.
Shifting the “see what sticks” mentality
Alex Kornilov, founder and CEO of advertising solutions supplier Betegy, believes that AI can bring about a shift in the “see what sticks” mentality that has dominated marketing in the gaming industry
Cutting his teeth in data and visualisation for sports, Kornilov soon realised that the sports betting industry had a much greater need for new ways to communicate complex data such as odds, stats and schedules.
After transitioning into the betting industry, he discovered that consumers in this space responded much better to personalised, dynamic advertising than the usual one-size-fits-all, static advertising. “Not just a little bit better – much better,” he explains.
“So you have a banner that says ‘bet now’, then you have a banner which says ‘bet now on the Champions League game Arsenal versus Barcelona’, this will convert much better.”
Betegy’s solution leverages data to create advertisements uniquely tailored to the player – from the sports they enjoy to the teams they follow – and then track the success of these advertisements and the value of the player over time.
AI is the answer
For this kind of deep segmentation, AI is the only solution, Kornilov says – especially in sports betting, where everything from the type of sport to the match schedule and the pre- and in-game odds must be taken into account.
“The depth of the product is so significant, so it does require a lot of data points to be connected to the system so that the ad you deliver to the audience is relevant and scalable,” he adds.
With AI in the driving seat, brands should be able to get a far more granular overview of how well their marketing is working for them. In future, it’s possible that the old joke – “Half the money I spend on advertising is wasted; the trouble is I don’t know which half” – will start to lose its relevance.
“Currently the acquisition strategy of every brand, especially in the United States, leads from spending money and seeing what sticks, which brings a lot of uncertainty at the brand level and no possibility to really tell what works and what doesn’t,” Kornilov explains.
“That’s why for example, last year there was a crisis on the acquisition side. If we acquire a player at 300 bucks and the lifetime value is ten, how long can we sustain this kind of aggressive growth?”
Painting a persona of gaming customers
As every operator knows, the challenge of growing revenues doesn’t begin and end with acquisition: retaining customers and increasing their lifetime value is crucial too.
This is where AI-driven personalisation can yield its best results, according to Epoxy’s Chris Reynolds.
With their insights tool, Epoxy can interrogate partners’ data to segment customers based on as many variables as the operator wants: their bet preferences, bet tolerances, their risk, their ability to become a VIP, their ability to churn, their favourite teams, favourite players, favourite bets, favourite parlays and so on.
This data can then be used in several ways to improve the client experience, increase retention, reduce churn and generally boost engagement, Reynolds says.
Negating the tyranny of choice
One particularly effective way to leverage this data is to create an experience that Epoxy has nicknamed “Betflix”: personalised websites and apps that are tailored to customers’ betting and gaming preferences, much like many popular streaming services and ecommerce sites.
There’s evidence that this type of interface can improve an app’s stickiness and reduce churn: in a recent Betting Hero study, 21% of New York bettors cited “ease of placing their favourite bets” as their main reason for returning to their favourite betting app.
Personalisation also helps counteract a phenomenon in human psychology known as choice overload, which describes the feeling of overwhelm people get when faced with too many options.
In ecommerce, choice overload might lead to people leaving an online store without making a purchase, while in betting, a player might feel overwhelmed and leave without ever placing a bet.
In theory, removing sports that players have no interest in and creating personalised recommendations will simplify a customer’s workflow and direct them to the bets they’re most likely to place. Put another way, a streamlined and personalised experience is all about negating the tyranny of choice.
Predicting the future
Although generative AI tools like ChatGPT have gained the most attention in the press in recent months, experts in the gambling industry say that predictive analytics, or predictive AI, has far more potential in this industry.
With a wealth of data on individual users’ preferences and past behaviour, it becomes possible to make relatively accurate predictions about their future behaviour, from the type of marketing that’s most effective to their betting habits and the likelihood of churn.
This type of cutting-edge analytics is already used in sports to track the health and performance of up-and-coming players, Betegy’s Kornilov reveals. Using key biomarkers gained from regular testing, sports coaches and teams can predict how likely a player is to suffer an injury, for example, or to start to show signs of decline.
Still early days for AI in gaming
Kornilov believes this kind of technology can also yield great results in the gambling industry, but says it’s still early days and “very much the Wild West”. Nevertheless, certain uses of predictive analytics can prove highly effective for operators.
According to the Betegy CEO, behavioural models can help identify a player who is likely to churn several weeks before they do so, giving CRM teams time to nurture and reactivate that customer relationship in a personalised way.
“All of this is very hard to segment by a human brain,” he explains. “You cannot create micro segments. That’s why you need to have certain tools and big data analytics and AI to make sure everything works in the best possible way.”
The two Epoxy co-founders also believe that predictive analytics and AI models should play a strong role in CRM in the gaming industry, giving operators an overview of the individual traits and behavioural patterns of a player over time.
“That allows an operator to much more effectively market to that user,” says Reynolds.
Tackling harm with AI
However, it’s not just in the realm of marketing and increasing retention where predictive models can come into play: they are also driving innovative and personalised solutions for reducing gambling-related harm.
One of the best examples of this use of AI is Mindway, the supplier of safer gambling software GameScanner, which leverages AI and cutting-edge neuroscience to identify players with a high potential to become problem gamblers in the future.
Mindway’s algorithms are trained using brain scans and the combined expertise of numerous psychologists, clinical psychiatrists and neuroscientists to create a scalable tool that can act as an early detection system.
According to testing and certification company Gaming Laboratories International (GLI), 99% percent of players who are manually identified as potential problem gamblers are also classed by GameScanner as either at-risk or potential problem gamblers, while at least 88.5% are classified as potential problem gamblers.
Not only does the tool give operators a wider overview of their pool of customers, but it also enables much more personalised interventions: for example, younger players could be reached more effectively via a push notification or text, while others may respond best to email.
“We have come a long way already when it comes to detection and monitoring of gambling behaviour and the next step we’re looking into is sophisticated tech support on the intervention part,” says Mindway’s Rasmus Kjaergaard.
“In other words: to empower the operators to do a lot more about the at risk and problem gambling behaviour our solutions detect.”
Tipping the scales
Although most people recognise how important AI and personalisation could be in the future, it seems that gambling industry is only at the beginning of this journey.
“I wouldn’t say the scales have quite tipped yet,” says Justin Le Brocque, manager of operations at Yolo Group. “Personalisation is absolutely an essential element; it is an expectation now rather than a bonus. But we are still able to deliver this manually, and often that is the best way to achieve the desired results.”
Nevertheless, Le Brocque acknowledges that “things are moving extremely quickly” and says Yolo is already implementing new AI-driven technology across the business in order to get ahead of the game.
“For now, we’re happy with the balance we’ve struck, but we expect that to evolve rapidly,” he says.
One of the next frontiers, according to industry experts, is the application of AI-driven personalisation in live betting to create augmented and integrated multimedia experiences.
Genius Sports’ AI vision
One of the latest products to explore this potential is BetVision, a solution from sports data specialists Genius Sports that pulls together live streaming, in-game betting and personalisation for a fully integrated sports betting experience.
Highlighting the significance of BetVision, Genius Sports product director Tom Holland describes it as “the amalgamation of everything we can do now”.
“When you’re starting to get events being identified by a computer in real time, you can then start to tie that back and try to create personal experiences for the user to really separate yourselves and create a unique experience for the consumer,” he says.
By gathering and processing live data from fast-paced sports events using AI, operators can create compelling narratives that appeal to fans’ emotions and drive engagement during the game, Holland adds.
“All the markets are essentially stories that I want to be a part of,” he explains. “What we’re trying to do is tie all of that together and make more granular stories that are focused on the narrative that you care about.”
Converging around a unified experience
Delivering this type of integrated and personalised experience is where many believe that AI could come into its own in the future.
“Fast forward three, five years, I think it’s hard to dispute that you’re going to see major technology companies converging around this unified experience which includes gamification, real money, betting, fantasy, all under one umbrella, which is really what the user wants,” Epoxy’s Angelides predicts.
“These silos are going to break down and the things that are logically going to converge are going to converge, particularly when it comes to media and sports betting.”
In four years, Angelides says, people will also be hard pressed to find a betting operator who isn’t using AI to personalise their product.
For Kornilov, this is precisely the reason why operators should investigate AI implementation as a strategic decision taken at the highest level.
“If you want to be competitive, you cannot allow yourself to not apply the most advanced way to solve your problems,” he says. “And the companies who implement this and can do it quickly and cheaply, they will win.”